Planejamento de vistorias usando robôs móveis autônomos e otimização pelo algoritmo de colônia de formigas

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

Robot-based inspection on sites with difficult access is a problem that becomes more common nowadays. Searching for solutions to this challenge, its observed new initiatives, either in orientated form in robot competitions like Robocup Rescue, or practical applications in companies that offer remoted operated vehicles to inspect pipes, galleries and further underground facilities. The natural evolution sequence for those services, as mentioned by several researches, is the automation of those tasks. With this automation, robots can accomplish, in an independent form, inspections of hard-access or dangerous places continually, without requiring the constant operation by a human being. This dissertation presents a set of heuristics of a multi-agent system based on positive reinforcements generated by a synergetic communication and negative reinforcements based on individual agent memory and modifications at the searching area using cultural algorithms to the automation of inspections. Beyond modifications on the collective intelligence, based on ant colonies behavior, its presented in this document the viability of the automation of this services, even on hard situations such as robots that require umbilical cables for data communication and power supply. Looking for improvement to the presented heuristics, detected by the of the algorithm convergence, this dissertation uses the Cultural Algorithms paradigm. Beyond adaptations of heuristics with this inspiration, the colony could reduce the searching space, allowing the optimization effort to be converted to the solution of a reduced part of the problem, this convergence could not be observed on the original algorithm.

ASSUNTO(S)

robôs móveis automação algoritmos automação algoritmos robótica engenharia de producao engenharia de producao robôs móveis robótica

ACESSO AO ARTIGO

Documentos Relacionados